Streamlining Resume Annotation with Named Entity Recognition

With the continuous development of natural language processing and its impact on artificial intelligence, various applications such as automated Q/A systems, Language translation, and Language generation and processing have garnered significant attention. One key aspect of NLP is text mining, encompassing technologies like text classification and NER (Named Entity Recognition). In real-world research and engineering projects, the availability of extensive and dependable text data sets is vital for ensuring the success of these tasks. However, in the IT, Higher Education, Legal, and Finance industries, the resumes flow in high volume but are very cluttered. Creating large-scale text data-sets of resume corpus has traditionally relied on manual sorting and labeling. This manual approach is resource-intensive and time- consuming, resulting in subjective variations due to diverse human perspectives. Consequently, these data sets may not fulfill the requirements of the actual industrial applications. To address these challenges, This paper presents a skillful approach for automatically annotating text data sets using NER (Named Entity Recognition) from the IT, Higher Education, Legal, and Finance sectors. This includes the combination of bi-LSTM and CRF, which can be a powerful assistant in labeling the data with multiple entities. We are also validating our generated corpus with CNN and RNN to ensure the quality of the corpus generated from the raw resumes.